Inferring Drug-Protein-Side Effect Relationships from Biomedical Text - MDPI

Page created by Sidney Fernandez
 
CONTINUE READING
Inferring Drug-Protein-Side Effect Relationships from Biomedical Text - MDPI
G C A T
T A C G
G C A T
           genes
Article
Inferring Drug-Protein–Side Effect Relationships
from Biomedical Text
Min Song 1, *, Seung Han Baek 2 , Go Eun Heo 1 and Jeong-Hoon Lee 3
 1   Department of Library and Information Science, Yonsei University, Seoul 03722, Korea;
     goeun.heo@yonasei.ac.kr
 2   Institute of Convergence, Yonsei University, Seoul 03722, Korea; seunghan.baek@yonsei.ac.kr
 3   Department of Creative IT Engineering, POSTECH, Pohang 37673, Korea; jhlee@dblab.postech.ac.kr
 *   Correspondence: min.song@yonsei.ac.kr
                                                                                                     
 Received: 24 January 2019; Accepted: 14 February 2019; Published: 19 February 2019                  

 Abstract: Background: Although there are many studies of drugs and their side effects, the
 underlying mechanisms of these side effects are not well understood. It is also difficult to
 understand the specific pathways between drugs and side effects. Objective: The present study
 seeks to construct putative paths between drugs and their side effects by applying text-mining
 techniques to free text of biomedical studies, and to develop ranking metrics that could identify the
 most-likely paths. Materials and Methods: We extracted three types of relationships—drug-protein,
 protein-protein, and protein–side effect—from biomedical texts by using text mining and
 predefined relation-extraction rules. Based on the extracted relationships, we constructed whole
 drug-protein–side effect paths. For each path, we calculated its ranking score by a new ranking
 function that combines corpus- and ontology-based semantic similarity as well as co-occurrence
 frequency. Results: We extracted 13 plausible biomedical paths connecting drugs and their side effects
 from cancer-related abstracts in the PubMed database. The top 20 paths were examined, and the
 proposed ranking function outperformed the other methods tested, including co-occurrence, COALS,
 and UMLS by P@5-P@20. In addition, we confirmed that the paths are novel hypotheses that are
 worth investigating further. Discussion: The risk of side effects has been an important issue for the
 US Food and Drug Administration (FDA). However, the causes and mechanisms of such side effects
 have not been fully elucidated. This study extends previous research on understanding drug side
 effects by using various techniques such as Named Entity Recognition (NER), Relation Extraction
 (RE), and semantic similarity. Conclusion: It is not easy to reveal the biomedical mechanisms of side
 effects due to a huge number of possible paths. However, we automatically generated predictable
 paths using the proposed approach, which could provide meaningful information to biomedical
 researchers to generate plausible hypotheses for the understanding of such mechanisms.

 Keywords: Biomedical Text Mining; semantic relatedness; Inference of Drug-Protein-Side
 Effect Relation

1. Introduction
      Minimizing drug side effects is a key focus of medical treatment as well as future drug
development. The side effects of drugs are well described in many clinical trials; however, these studies
simply report on a given drug’s side effects and not attempt to explain the cause [1–3]. Because many
factors are likely to connect a drug and a particular side effect, time and effort is required to discover
the relationship between the two. Text mining approaches have been proposed as way to examine
this relationship.
      While many text mining studies have extracted overall relationships between drugs and side
effects, they could not suggest the series of mechanistic steps from a drug to a particular side effect [4–8].

Genes 2019, 10, 159; doi:10.3390/genes10020159                                   www.mdpi.com/journal/genes
Inferring Drug-Protein-Side Effect Relationships from Biomedical Text - MDPI
Genes 2019, 10, 159                                                                                2 of 17

Therefore, we attempted to use text mining to identify concrete paths from a given drug and side
effect. That is, if a drug and side effect are given as start and end points, respectively, the series of
mediators (i.e., proteins) are extracted automatically from the text and connected to form a proposed
drug-protein–side effect path.
      Although there are public databases of pathways such as the KEGG (Kyoto Encyclopedia of
Genes and Genomes) database, these databases have limited information and are not specialized
for specific drug side effects. Using the Swanson ABC model as our foundation, we extracted the
relation between entities and connected them by using the text mining method to study the extracted
paths. There are several studies utilizing text mining to discover relationships between drugs and
their side effects. Sohn et al. [9] proposed the decision tree approach, a machine learning algorithm,
to extract sentences that contain drug and side effect pairs. To build feature sets, they employed
side effect keyword features and pattern matching rules. Zhang et al. [10] developed a similarity
driven matrix factorization method to predict the drug and side effect associations. The primary goal
of their approach was to estimate the drug-side effect association relationship by identifying latent
features for drugs and side effects to compute drug feature-based similarities and disease semantic
similarity. Pauwels et al. [11] proposed a sparse canonical correlation analysis (SCCA) to predict
potential side-effect profiles of candidate drugs based on their chemical structures. The primary goal
of their approach was to extract correlated sets of chemical substructures and side-effects. Previous
studies on the relationship between drugs and side effects primarily focused on the direct link between
these two types of entities. Unlike these studies, the proposed approach is based on the indirect link
between drugs and side effects via proteins.
      In the present study, we used advanced text-mining techniques to extract relations between
entities related to cancer drug research and to form a network with which to identify the chain
reaction that occurs in the body when a drug is given [12–18]. We derive this network based from
the interactions of entities suggested in scientific papers [19–24]. It is an accumulated and integrated
network that connects fragmentary relations between drug-protein, protein-protein, and protein–side
effect, thereby connecting drugs to their side effects [25–28].
      Based on this composed knowledge network, we analyzed paths between drugs and side effects
and ranked them using a new ranking function that combines semantic similarity scores and the
frequency of entity pair co-occurrence to suggest meaningful paths. We compared our proposed
ranking algorithm with other algorithms such as Correlated Occurrence Analogue to Lexical Semantics
(COALS), and ontology-based algorithm (UMLS; Unified Medical Language System), and showed
that the proposed algorithm outperforms the other three in terms of precision of K measurements.
      Analyzing 20 identified paths, we were able to adjudge 65% of them as biologically plausible.
Although our research does not show perfect performance, it can help to identify the link between
drugs and side effects, and serve as a foundation for further research. Moreover, our method can be
applied to other types of paths, thereby revealing previously unknown links among biological entities
such as diseases and genes.

2. Materials and Methods
      To investigate possible drug-protein–side effect paths, we propose a three-step framework:
(1) pre-processing, (2) named entity recognition (NER) and relation extraction, and (3) path analysis
(illustrated in Figure 1). Most cancer drugs target specific paths, which are composed of protein
interactions. We therefore established proteins as our middle node. The first component is
pre-processing, including parsing XML records that were downloaded manually from the PubMed
search engine. During the parsing of records, we extracted important metadata such as PMID,
title, and abstract. Once parsing was done, an abstract of the record was split into sentences by a
sentence boundary detection algorithm; we used the algorithm provided in Stanford CoreNLP [29].
The second component has something to with NER and relation extraction. For NER, we employed
a dictionary-based entity extraction. Since an entity is either a single word or a phrase, we needed
Inferring Drug-Protein-Side Effect Relationships from Biomedical Text - MDPI
Genes 2018, 9, x FOR PEER REVIEW                                                                                            3 of 17
  Genes 2019, 10, 159                                                                                                           3 of 17

second component has something to with NER and relation extraction. For NER, we employed a
dictionary-based       entity extraction.
   to tokenize sentences          by N-gram.   Since    an entity
                                                   N-gram           is either
                                                              denotes            a single word
                                                                          a contiguous              or a of
                                                                                              sequence    phrase,    we needed
                                                                                                              n tokens.             to
                                                                                                                           Once entity
tokenize   sentences       by   N-gram.      N-gram      denotes    a  contiguous       sequence
   extraction is done, extracted entities and a sentence are fed into the relation extraction module.of  n  tokens.     Once   entity
extraction   is done,
   In relation            extracted
                   extraction,          entities and (POS)
                                   Part-Of-Speech        a sentence     are fed
                                                                  tagging     wasinto     the relation
                                                                                    applied               extractiontomodule.
                                                                                                to the sentence                     In
                                                                                                                          detect verbs.
relation  extraction,      Part-Of-Speech       (POS)    tagging    was    applied    to  the  sentence
   Since not every verb is meaningful in the bio-medical domain, we adopted the list of biomedical         to detect   verbs.   Since
notverbs
     everyprovided
             verb is meaningful
                         in [10]. Theinthird the component
                                                   bio-medical was  domain,
                                                                          path we    adopted
                                                                                 analysis.        theextracted
                                                                                               With    list of biomedical
                                                                                                                   entities and verbs
                                                                                                                                    their
provided    in  [10].  The    third  component       was   path   analysis.    With    extracted
   relations, built a graph and generated k-depth paths. Once paths were generated, we computed    entities   and   their  relations,
built  a graph
   semantic         and generated
                relatedness      scores k-depth
                                          of any givenpaths.
                                                           pairOnce    paths linked
                                                                 of entities    were generated,
                                                                                         on the graph. we Tocomputed
                                                                                                               this end, semantic
                                                                                                                            we utilized
relatedness     scores    of  any   given    pair   of entities   linked    on   the  graph.
   two different sources: UMLS and the collected dataset a.k.a. corpus. UMLS was used to computeTo  this  end,    we  utilized    two
different   sources:between
   the similarity        UMLS two  and entities
                                          the collected
                                                    based ondataset    a.k.a. corpus.
                                                                the concept                UMLS
                                                                                 hierarchical       was used
                                                                                                 structure        to compute
                                                                                                              of UMLS.             the
                                                                                                                            The corpus
similarity   between      two   entities   based   on  the  concept    hierarchical      structure   of
   was used to build a corpus-based semantic relatedness model so that we could compute the similarity  UMLS.     The   corpus    was
used   to build
   of two          a corpus-based
            entities   found in thesemantic
                                          model. To  relatedness
                                                        make it easy model     so that we
                                                                          to examine      thecould   compute
                                                                                               results  reported  thebysimilarity   of
                                                                                                                         the proposed
two   entities   found    in   the  model.    To   make    it easy   to examine      the   results
   approach and as a utility function for the present paper, we developed a simple JSP (Java Server reported     by  the   proposed
approach     and as
   Page)-based         a utility
                     web    serverfunction
                                     that allowedfor theus present    paper,
                                                            to navigate      thewetopdeveloped
                                                                                        350 paths.a Thesimple
                                                                                                            webJSP     (JavaisServer
                                                                                                                   server      publicly
Page)-based      web    server    that  allowed      us to  navigate     the   top  350   paths.   The
   accessible at http://informatics.yonsei.ac.kr:8080/drug_se/searchDrug.jsp. The web server provides    web   server    is  publicly
accessible
   a functionat http://informatics.yonsei.ac.kr:8080/drug_se/searchDrug.jsp.
                  of search by drug. The matched search result displays paths starting              The web     server
                                                                                                              with   drug,provides   a
                                                                                                                             along with
function   of  search    by   drug.  The    matched     search   result   displays     paths   starting
   a PubMed ID that takes the reader to the PubMed record page. The results page also provides a link to  with   drug,   along   with
a PubMed
   the PubChemID that    takes
                      site  for athe  reader todrug.
                                   particular      the PubMed       record
                                                         In addition,     eachpage.
                                                                                 pathThe     resultsdown
                                                                                        is broken     page also
                                                                                                             into aprovides     a link
                                                                                                                     set of pairs   such
to as
   thedrug
        PubChem
             and protein, protein and disease, etc. It is also displayed that a sentence of a PubMedpairs
                      site   for  a particular     drug.  In  addition,     each   path   is  broken   down     into  a  set of   record
such   as drug
   contains    theand   protein,pair
                    particular      protein
                                         in theand   disease, etc. It is also displayed that a sentence of a PubMed
                                                  path.
record contains the particular pair in the path.

      Figure 1. System overview.
                                                     Figure 1. System overview.

2.1.2.1.
     Data  Source
         Data     andand
              Source  Pre-processing
                         Pre-Processing
     WeWe first
            firstselect
                  selectaa specific  domainrelated
                           specific domain      relatedtoto  cancer
                                                          cancer  andand    retrieve
                                                                       retrieve       relevant
                                                                                 relevant        data PubMed
                                                                                          data from     from PubMed
                                                                                                                  database
database  (http://www.ncbi.nlm.nih.gov/pubmed).
   (http://www.ncbi.nlm.nih.gov/pubmed).                     By consulting
                                                        By consulting    withwith    several
                                                                                several       domain
                                                                                         domain         experts
                                                                                                   experts        on query
                                                                                                             on the  the
query   terms,    by   researching    cancer-related     bio  articles,  and    by  referring   to  the
   terms, by researching cancer-related bio articles, and by referring to the cancer gene information in  cancer   gene
information
   the KEGGin     the KEGG
                database,   wedatabase,
                                selected 37wekeywords
                                               selected 37related
                                                            keywords     related
                                                                   to cancer   andtoextracted
                                                                                     cancer and   extracted
                                                                                                abstracts     abstracts
                                                                                                           from   PubMed
from  PubMed
   database  only database   only if they
                    if they contained        contained
                                         these  keywords.these   keywords.
                                                             In this way, weIncollected
                                                                                  this way,   we collected
                                                                                          a total             a total
                                                                                                   of 2,379,349       of
                                                                                                                  abstracts;
2,379,349  abstracts;   Figure  2 shows    the
   Figure 2 shows the search query terms used. search  query   terms  used.
     In the
        In pre-processing
            the pre-processingstage, we   extracted
                                      stage,        the PubMed
                                               we extracted        ID,PubMed
                                                                 the   title, and abstract from
                                                                                   ID, title,    the abstract
                                                                                               and    PubMed records
                                                                                                                 from the
(which  are inrecords
   PubMed       XML form)     using
                         (which    areSAX-based
                                        in XML XML form)parsing
                                                           using module.
                                                                    SAX-based  We XML
                                                                                   also used  the Stanford
                                                                                          parsing    module.CoreNLPWe also
text-mining  tool,  which  includes   sentence   segmentation,    POS  tagging,   and  lemmatization
   used the Stanford CoreNLP text-mining tool, which includes sentence segmentation, POS tagging,       [29].
   and lemmatization [29].
Inferring Drug-Protein-Side Effect Relationships from Biomedical Text - MDPI
Genes 2019, 10, 159                                                                                              4 of 17
Genes 2018, 9, x FOR PEER REVIEW                                                                          4 of 17

     Figure 2. Search query.
                                                  Figure 2. Search query.
2.2.2.2.
     Named    Entity
          Named      Recognition
                  Entity Recognition
      Next we identify drug, protein, and side effect entities that mentioned in the text and exist in the
           Next we identify drug, protein, and side effect entities that mentioned in the text and exist in the
databases using dictionary-based Named Entity Extraction [30].
    databases using dictionary-based Named Entity Extraction [30].
      There are several open source NER tools such as MetaMap, Banner, ABNER, and LingPipe.
           There are several open source NER tools such as MetaMap, Banner, ABNER, and LingPipe.
Although machine-learning approaches including Banner, ABNER, and LingPipe are prevalent due
    Although machine-learning approaches including Banner, ABNER, and LingPipe are prevalent due to
to their efficiency, they often suffer from poor performance when applied to real world scenarios.
    their
Thus,    inefficiency,
            our study, they often suffer
                         because         from poor
                                   performance      performance
                                                 accuracy          when important
                                                            is far more  applied to than
                                                                                      real world    scenarios.
                                                                                             efficiency, we Thus,
    in our study,
constructed          because
               an entity      performance
                         dictionary         accuracy
                                    to exactly       is far more
                                               match entities withimportant   than the
                                                                   sentences from   efficiency,
                                                                                        full text.we constructed an
    entity dictionary to exactly match entities with sentences from the full text.
2.2.1. Building the entity name dictionary
    2.2.1. Building the Entity Name Dictionary
      We constructed the entity name dictionary using consolidated entity names from three
publicallyWeavailable
              constructed    the entity
                         databases:       name dictionary
                                       Drugbank     Version using      consolidated entity namesfor
                                                               4.1 (http://www.drugbank.ca/)        from
                                                                                                       drugthree publically
                                                                                                              type
entities, Uniprot (http://www.uniprot.org/) for human protein type entities, and SIDER Uniprot
    available  databases:    Drugbank     Version   4.1  (http://www.drugbank.ca/)         for drug type  entities,
    (http://www.uniprot.org/)
(http://sideeffects.embl.de/)          for effect
                                 for side  human    protein
                                                  entities.   type entities,
                                                            Finally,          and SIDER
                                                                     we formulated          (http://sideeffects.embl.de/)
                                                                                      the expanded   dictionary to
    for side
include       effect entities.
         synonyms               Finally,
                      of each entry   namewederived
                                               formulated
                                                        fromthe   expanded
                                                              these  databases.dictionary to include synonyms of each
    entry  name
      In the  casederived    from these
                    of proteins,           databases.
                                    multiple    synonyms exist for a singular protein, making synonym
processing    essential.
          In the          In our research,
                 case of proteins,   multiplewe     used theexist
                                                synonyms        synonyms     that are
                                                                    for a singular     provided
                                                                                   protein,       by synonym
                                                                                             making  the UniProtprocessing
database   to construct
    essential.           our synonym
                In our research,   we useddictionary   for proteins
                                              the synonyms            as well
                                                                that are      as their
                                                                          provided  bysynonymous     gene names.
                                                                                       the UniProt database    to construct
    our synonym dictionary for proteins as well as their synonymous gene names.
2.2.2. Recognition of entity name
    2.2.2. Recognition of Entity Name
      We used N-gram matching to recognize the entities in sentences we extracted. Sentences are first
split andWe then
              usedN-gram
                     N-gram tokenized
                               matching  bytoApache    Lucence
                                               recognize          (http://lucene.apache.org).
                                                            the entities                        N-gramSentences
                                                                          in sentences we extracted.     has been are first
performed
    split andbythen
                 1-, 2-, 3-gram,
                     N-gram        and thesebysplit
                                tokenized             tokens
                                                  Apache       (token n-gram)
                                                             Lucence             are identified within N-gram
                                                                        (http://lucene.apache.org).     the namehas been
entity
    performed by 1-, 2-, 3-gram, and these split tokens (token n-gram) are identified withinbytheexact
        through    three   different  entity   dictionaries.   NER    for  each n-gram    is conducted         name entity
matching.
    through three different entity dictionaries. NER for each n-gram is conducted by exact matching.

2.3.2.3.
     Relation Extraction
         Relation Extraction
    Before   merging
         Before         the the
                   merging    three multi-type
                                  three         entity
                                        multi-type     paths,
                                                    entity     we first
                                                           paths,       focused
                                                                   we first      on the
                                                                            focused     relationship
                                                                                     on the           between
                                                                                            relationship  between two
twoentities.
    entities.IfIfaaverb
                    verbisislocated
                             locatedbetween
                                      betweentwo
                                               twoentities
                                                    entitiesin
                                                             inaasentence
                                                                  sentenceand
                                                                            andcorresponds
                                                                                correspondsto tothe
                                                                                                 therules
                                                                                                     rulesthat
                                                                                                           thatwe have
we have made (see below), then we extracted the two entities assuming that they were related, along
   made (see below), then we extracted the two entities assuming that they were related, along with the
with the verb. This relation-extraction module is executed to construct three different relations—
   verb. This relation-extraction module is executed to construct three different relations—drug-protein,
drug-protein, protein-protein, and protein–side effect—which are then merged to create the entire
   protein-protein, and protein–side effect—which are then merged to create the entire paths of drug-
paths of drug- protein–side effect.
   protein–side effect.
Genes 2019, 10, 159                                                                                                             5 of 17

 GenesTo  extract
       2018,        thePEER
             9, x FOR    drug-protein,
                               REVIEW          protein-protein, and protein–side effect relations from sentences5 that              of 17
include annotated entities, we prepared and applied these rules to the extraction module with Stanford
CoreNLPTo extract
             toolkit.the drug-protein, protein-protein, and protein–side effect relations from sentences
 that include annotated entities, we prepared and applied these rules to the extraction module with
1.Stanford     rule: When
       BasicCoreNLP       toolkit. the sentence structure is presented as entity–verb–entity, we assume that
 1. theBasictworule:entities
                       When have          a relationship.
                                 the sentence       structure is Topresented
                                                                      identify the      main verb in a sentence,
                                                                                   as entity–verb–entity,       we assume    wethatused
                                                                                                                                      the
      part-of-speech       tagging.
       two entities have a relationship. To identify the main verb in a sentence, we used part-of-speech
2.     tagging. detection: When a sentence includes negation dependency relation “neg” in the
       Negation
 2. dependency
       Negation detection:
                        parse treeWhen    or whena sentence
                                                      a token includes
                                                                 matches with negation      dependency
                                                                                     the negative       wordrelation
                                                                                                               such as“neg”
                                                                                                                         “hardly” in the
                                                                                                                                       or
       dependency
      “scarcely,”     weparse
                          classifiedtree them
                                          or when      a token matches
                                                  as negative.                with all
                                                                  We classified       theother
                                                                                           negative      wordassuch
                                                                                                 sentences            as “hardly” or
                                                                                                                  non-negative.
3.     “scarcely,”
       Relation        we classified
                    keyword:        Afterthem      as negative.
                                            implementing            We classifiedtagging,
                                                               part-of-speech          all otherwe sentences    as non-negative.
                                                                                                      filtered whether     the verb is
 3. included
       Relation in  keyword:         After   implementing       part-of-speech       tagging,
                       the list of 389 verbs that are commonly used in the biomedical domain, as  we   filtered whether     thedefined
                                                                                                                                 verb is
       included
       by  previous in the
                        work  list[31].
                                     of 389
                                          Weverbs     that are commonly
                                               also manually       classified theused   in the
                                                                                      verb   listsbiomedical    domain, asincrease
                                                                                                    into two categories:        defined
       by previousenhance,
      (accelerate,      work [31].        We alsoactivate,
                                      stimulate,      manually      classified
                                                                 etc.) or decreasethe verb    lists reduce,
                                                                                        (inhibit,    into twoabolish,
                                                                                                                categories:    increase
                                                                                                                        silence,    etc.)
       (accelerate,
       based   on theenhance,
                        bio-verb stimulate,
                                       list.         activate, etc.) or decrease (inhibit, reduce, abolish, silence, etc.)
4.     based on the
       Distance          bio-verb
                    between             list. The distance, or number of words, between entities linked by a
                                   entities:
 4. bio-verb
       Distance(relation
                     between keyword)
                                   entities: The     distance,
                                                 is an   importantor number
                                                                        factor in of words,
                                                                                     deciding  between
                                                                                                   whether  entities linked
                                                                                                               the two        by a bio-
                                                                                                                         entities    are
       verb related.
       truly  (relationWe  keyword)
                                chose a is      an important
                                            window                factor
                                                        size of six        in deciding
                                                                      words,                whether the
                                                                                 and we extracted         twotwo   entities
                                                                                                               entities and are    truly
                                                                                                                               relation
       related. We
       keywords         chose
                     only   if they a window
                                        are includedsize within
                                                          of six the
                                                                   words,
                                                                        windowand size.
                                                                                     we extracted two entities and relation
5.     keywords       only   if they    are   included    within
       Direction: Direction can be determined by the tense (active  the  window      size.or passive) using Stanford CoreNLP
 5. part-of-speech
       Direction: Directionparsing    can  be dependency
                                         and    determined by       the tense
                                                                parsing          (active
                                                                           results.    Whenor passive)
                                                                                                “auxpass”  using Stanford CoreNLP
                                                                                                              dependency       relation
       part-of-speech       parsing       and   dependency       parsing    results.   When      “auxpass”
       appears in a given sentence or when the sentence corresponds to the rules such as auxiliary             dependency       relation
                                                                                                                                   verb
      +appears     in a given
         past participle       andsentence
                                      be verbor+when       the sentence
                                                     past participle,     wecorresponds
                                                                               consideredto       the rules
                                                                                               them           such
                                                                                                        passive     as auxiliary
                                                                                                                  voice;  the restverb
                                                                                                                                     we
       considered active. If a relation keyword is passive, the direction of verb is reversed: “entity Awe
       + past   participle      and    be   verb   + past   participle,   we    considered      them     passive  voice;   the  rest   is
       considered
       activated    byactive.
                        entity If  B”a means
                                        relation   keyword
                                                 “entity        is passive,
                                                           B activates         theA.”
                                                                          entity    direction of verb is reversed: “entity A is
       activated by entity B” means “entity B activates entity A.”
2.4. Path Detection on Drug-Protein–Side Effect
 2.4. Path Detection on Drug-Protein–Side Effect
      For path analysis, we combined the three different entity relationships generated in the
       For path analysis,
aforementioned        stages (Figure  we combined
                                            3). Proteinsthe thatthree
                                                                  appear different
                                                                            in both entity
                                                                                       pairs ofrelationships
                                                                                                  drug-protein generated         in the
                                                                                                                   and protein–side
 aforementioned        stages    (Figure     3). Proteins
effect work as a bridge that connects drug and side-effect.  that  appear    in both   pairs   of  drug-protein    and   protein–side
 effect work as a bridge that connects drug and side-effect.

         Figure3.3.Conceptual
        Figure     Conceptualmodel
                              modelof
                                    ofdrug-protein–side
                                       drug-protein–sideeffect
                                                         effectpath.
                                                                path.D—drug,
                                                                      D - drug,P—protein,
                                                                               P – protein,S—side
                                                                                            S – sideeffect.
                                                                                                     effect.
2.4.1. Pair Generation of Drug-Protein
 2.4.1. Pair generation of drug-protein
      When we merged the heterogeneous path of the three entities, we first selected 50 drug entities
       When we merged the heterogeneous path of the three entities, we first selected 50 drug entities
related to cancer that appear in the DrugBank database (listed in Table 1). We then detected the
 related to cancer that appear in the DrugBank database (listed in Table 1). We then detected the drug-
drug-protein pairs related to these 50 cancer drugs.
 protein pairs related to these 50 cancer drugs.
Genes 2019, 10, 159                                                                                    6 of 17

                                       Table 1. List of selected drugs.

         Drug Name           Drug Name            Drug Name               Drug Name       Drug Name
       amifostine             Cetrorelix             erlotinib              Letrozole        sorafenib
    aminoglutethimide       chlorambucil          exemestane               Leucovorin       tamoxifen
        amsacrine              cisplatin          fludarabine              Lomustine     temozolomide
       anagrelide             cladribine            flutamide             methotrexate    temsirolimus
       anastrozole           clofarabine             gefitinib              Mitotane          thiotepa
       bexarotene            dacarbazine          gemcitabine               Nilotinib       topotecan
      bicalutamide          daunorubicin           idarubicin              nilutamide      toremifene
         busulfan             degarelix            ifosfamide             ondansetron       vincristine
      capecitabine            docetaxel            irinotecan               paclitaxel     vinorelbine
       carboplatin           doxorubicin          ixabepilone             procarbazine     bortezomib

2.4.2. K-Protein Depth Path
     The path can differ depending on the number of protein entities, as they are bridges connecting
drug and side effect. The number of protein entities (k) can be presented as the depth of path (k-protein).
Altering k from 1 to 3, we investigated not only the direct paths of drug-protein–side effect, but also
the chain reaction of path between proteins. For example, depth 1 is drug-protein1–side effect and
depth 3 is drug-protein1-protein2-protein3–side effect.

2.4.3. Path Ranking Algorithm
      To rank the integrated paths of drug-protein–side effect, we combined three metrics: i) the lexical
co-occurrence–based semantic similarity score COALS [25,26]; ii) the frequency score of the pair’s
co-occurrence; iii) the knowledge-based semantic similarity score derived from UMLS [27,28].
      First, COALS was used to estimate the similarity between two words with a given text corpus.
Basically, it was calculated by the correlation of two co-occurrence vectors both for normalization and
for measuring vector similarity. Second, the frequency-weighted score was calculated as the number of
occurrences of a pair on k-depth path. From the score of maximum frequency, we calculated the local
frequency, based on the specific side effect. Third, we used knowledge-based semantic similarity score
determined from a manually developed knowledgebase like UMLS, which gives the general biological
relationship between two given entities.
      As a linear combination of the three metrics, we ranked the path (α = 0.3, β = 0.4, γ = 0.3).

                                                   LocalFreq(vi ,v j ,SEk )                             
  Entity Pair Weight vi , v j = α·COALS vi , v j + β· MaxLocalFreq            + γ·U MLS_Semantic vi , v j
                                                                           SEk

                                                n −1
                                                        Entity Pair Weight(vi , vi+1 )
                         Path ranking score =    ∑                  n−1
                                                 i =0

where vi : Certain entity in given path (drug-protein–side effect sequence); COALSvi ,v j : COALS
                                                                                                          
semantic similarity between vi and v j , where vi and v j are extracted entities; LocalFreq vi , v j , SEk :
Frequency between vi and v j among all path of specific side effect; MaxLocalFreqSEk : Maximum
                                                                                   
frequency score of specific side effect among LocalFreq; U MLS_Semantic vi , v j : Knowledge-based
semantic similarity between vi and v j .

3. Results
     After extracting the relationship between entities from free text using the proposed
relation-extraction method, we constructed integrated networks consisting of drug-protein,
protein-protein, and protein–side effect pairs. We then identified a series of paths from drug to
side effect. Paths were ranked in descending order by the proposed weight function.
Genes 2019, 10, 159                                                                                                    7 of 17

3.1. Construction of Drug-SE (Side Effect) Path

3.1.1. Extraction of Relation Pairs from Text
     The numbers of entity relation pairs extracted from our dataset are in Table 2. There are 1430 kinds
of cancer drugs that appeared from the drug-protein pairs extracted from the text, and 2156 types of
cancer drug side effects from the protein–side effect pairs. This is 18.47% of all drugs registered in the
DrugBank and 33.78% of all the side effects in the dictionary we built based on SIDER.

                      Table 2. Number of pairs, source, and target depend on each entity relation.

          Coverage              Entity Relation       # of Pairs          # of Unique Source        # of Unique Target
                                 drug-protein               32,307                 1430                    2709
          All drugs             protein-protein            146,912                 6125                    5904
                               protein–side effect         170,112                 6222                    2156
                                 drug-protein               2622                    50                     1055
      50 cancer drugs
                               protein–side effect         41,415                  5313                    996

     When limiting our list of target drugs to the previously selected 50 cancer-related drugs and to
the relevant side effects, all 50 drugs were listed in drug-protein pairs, while only some of the side
effects were extracted from protein–side effect pairs. Related to the selected 50 drugs, all 1988 side
effects appeared in SIDER, but only 996 cases were extracted from the text. In other words, we did not
extract all the side effects of cancer-related drugs listed in SIDER for the following two reasons: First,
350 side effects were not mentioned in the collected dataset; Second, 543 side effects were not extracted
by the dictionary-based side effect extraction, suggesting that our analysis does not cover all cases.

3.1.2. Detection of Drug-Protein–Side Effect Path
     To understand the path from drug to side effect, we created paths made of three types of entities by
combining drug-protein pairs with protein–side effect pairs. Then, we categorized the paths by k, the
number of protein existing between drug and side effect (denoted as k-protein depth). The suggested
prioritizing algorithm ranks each path according to the proposed weight function, and we looked into
the top 250 paths with k values of 1, 2, and 3. The number of unique drugs, proteins, and side effects
are shown in Table 3.

                              Table 3. Number of drug, protein, and SE in top 250 paths.

                          Path Type                          # of Paths    # of Drugs     # of Proteins   # of Side Effects
            drug-protein–side effect (depth-1)                  250           28              131                 17
       drug-protein1-protein2–side effect (depth-2)             250           34              126                 32
   drug-protein1-protein2-protein3–side effect (depth-3)        250           39              222                 78

     The extracted paths are shown in the form of drug-protein–side effect. However, these include
the paths of the drugs that elicited a treatment response, and so include both intended effect and
side effect. For example, there were relatively many tumor or cancer cases in side effect section,
when a protein and “tumor” were connected by verbs such as “reduce.” As k increases, the number of
drugs and side effects increase as well. In drug-protein–side effect paths, 17 cases of side effects were
shown redundantly, 32 redundant cases of side effects in drug-protein1-protein2–side effect paths,
78 redundant cases of side effects in drug-protein1-protein2-protein3–side effect paths. The number
of drugs also increased by 28 in drug-protein–side effect paths, by 34 in drug-protein1-protein2–side
effect paths, and by 39 in drug-protein1-protein2-protein3–side effect paths.
Genes 2019, 10, 159                                                                                                                               8 of 17

3.1.3. Selection of Significant SE Path
     The actual side effects and their paths were difficult to track because the prior top 250 paths
included both the intended effects of the drugs and their side effects. Therefore, we classified the paths
into effects and side effects by considering paths ending with the nodes such as tumor or cancer as
effects and excluded it from our extracted paths.
     Among the remaining side effect–specific paths, we then selected only significant paths by
inspecting extracted verbs. A path was considered significant only if the verbs represent a change to
other bio entities. Table 4 shows the final top 20 significant paths ranked according to our ranking
function described in Section 2.4.3. We evaluate the top 20 ranked paths. Our work provides a
hypothesis as a starting point for new biological research with relatively little time and effort.

                                                      Table 4. List of top 20 paths.
  No       Drug            Verb1          Protein      Verb2        Protein       Verb3        Protein         Verb4             Side Effect
   1    anastrozole       observe           ar        decrease      muc5ac         use       polymerase        suggest          acute hepatitis
   2    anastrozole        differ          age        associate     muc5ac         use       polymerase        suggest          acute hepatitis
                                                                                 inhibit,
   3     irinotecan        inhibit         p38         inhibit        g17                      gastrin        associate           dyspepsia
                                                                                neutralize
                                                                                 inhibit,
   4     tamoxifen     abolish, delete     p38         inhibit        g17                      gastrin        associate           dyspepsia
                                                                                neutralize
                                                                                 inhibit,
   5    doxorubicin    induce, activate    p38         inhibit        g17                      gastrin        associate           dyspepsia
                                                                                neutralize
                           reduce,                                               inhibit,
   6     nilotinib                         p38         inhibit        g17                      gastrin        associate           dyspepsia
                          increase                                              neutralize
                                                                                 inhibit,
   7     sorafenib      inhibit, block     p38         inhibit        g17                      gastrin        associate           dyspepsia
                                                                                neutralize
                           inhibit,                                              inhibit,
   8    bortezomib                         p38         inhibit        g17                      gastrin        associate           dyspepsia
                          decrease                                              neutralize
                                          protein
   9    bortezomib         induce                   phosphorylate    p150         occur          cd5           present      glomerulonephropathy
                                          kinase
                           reduce,
   10    cetrorelix                         egf       decrease        p15       increase       smad4           interact          septal defect
                          decrease
   11    cetrorelix         inhibit        pcna       conserve        p15       increase       smad4           interact          septal defect
                           induce,
   12   bortezomib                         p53        inactivate     p150         occur          cd5           present      glomerulonephropathy
                          stimulate
                                                                                 identify,
   13   gemcitabine       increase          il-2      stimulate       gls                    glutaminase      catalyze     nervous system disorders
                                                                                  serve
                          decrease,
   14   doxorubicin                         il-6       interact      clec-2       serve      podoplanin       accelerate         leukoplakia
                          enhance
                           reduce,                    enhance,
   15    nilotinib                         p38                       mao-a      increase        ssao           predict     intracranial hemorrhage
                          increase                     inhibit
                           induce,                    increase,
   16    cisplatin                         tnf-α                      spo          use         lipase           occur        hypophosphatemia
                          increase                    decrease
                                                                                 result,
   17   methotrexate       inhibit          lp          result        nrl                        rod            lead             nocardiosis
                                                                                become
                          induce,                                               enhance,
   18   chlorambucil                       p53        promote       l3mbtl1                  erythropoietin     exert        ureteral obstruction
                        up-regulate                                             decrease
                        up-regulate,
   19   methotrexate                        ts         encode       kinesin-2    reduce          rod            lead             nocardiosis
                          inhibit
                         separate,
   20   methotrexate                        cr        increase      cofilin-1   correlate        rod            lead             nocardiosis
                          observe

3.2. Verification of Path
     To measure the performance of the proposed method, we examined the top 20 paths to determine
whether they were conceptually plausible or not. Because the results covered a variety of biological
entities, we also retrieved the sentences deriving entity pair of the path from PubMed in order to
prevent the misunderstanding of such a path.
     We analyzed the paths and classified them into three types. Type 1 paths involve entities that are
related and verbs that are also correctly connected. Type 2 paths involve entities that are related but
verbs that are uncertain. And type 3 involves entities of uncertain relation as well as uncertain verbs.

3.2.1. Type Description
      We were capable of interpreting type 1 and type 2 paths as shown in Table 5. Although the verbs
that connect the entities are not certain in case of type 2, we were able to connect the entities through a
literature analysis showing that the entities are highly related. Through our literature analysis, we were
Genes 2019, 10, 159                                                                                 9 of 17

able to verify 65% of the 20 paths as plausible. The verified 65% supports our suggested path between
drug and side effect. The result of each path is given in Table 5.

                              Table 5. Biological analysis result (top 20 path).

          Path No.        Drug                 Side-Effect               Type1     Type2   Type3
               1.      anastrozole            acute hepatitis                       O
               2.      anastrozole            acute hepatitis                               O
               3.       irinotecan               dyspepsia                 O
               4.       tamoxifen                dyspepsia                 O
               5.      doxorubicin               dyspepsia                 O
               6.        nilotinib               dyspepsia                 O
               7.        sorafenib               dyspepsia                 O
               8.      bortezomib                dyspepsia                 O
               9.      bortezomib         glomerulonephropathy                              O
              10.        cetrorelix            septal defect                        O
              11.        cetrorelix            septal defect                        O
              12.      bortezomib         glomerulonephropathy                              O
              13.     gemcitabine       nervous system disorders                            O
              14.      doxorubicin              leukoplakia                         O
              15.        nilotinib       intracranial hemorrhage                    O
              16.        cisplatin          hypophosphatemia                        O
              17.     methotrexate              nocardiosis                                 O
              18.     chlorambucil          ureteral obstruction                    O
              19.     methotrexate              nocardiosis                                 O
              20.     methotrexate              nocardiosis                                 O

     There are some cases in which same pair of drug and side effect appear twice, but are connected
by different proteins. For example, the top two ranked paths both link the cancer drug anastrozole to
the side effect of acute hepatitis.

•     Type 1: When the entities are related and the verbs that describe the relations are also
      correctly connected.
•     Type 2: When the entities are related but the verbs that describe the relation are not certain.
•     Type 3: When the relation of entities are not certain and the verbs that describe the relation are
      also not certain.

3.2.2. Comparison Ranking Functions
     For the selected 20 paths, we compared the performance of the proposed method and existing
measures: (1) the average betweenness degree using the co-occurrence frequency of two entities, (2) the
semantic similarity obtained by utilizing the COALS algorithm, (3) the semantic similarity for the
biomedical domain within the framework of UMLS. We judged that the information corresponding
to type 1 is a correct path, and that corresponding to type 3 is an incorrect path. We calculated top
n-ranked paths and determined how many correct paths exist. We compared the results between the
ranking function we proposed and the other functions. We employed precision at k (P@k), a common
measurement in the field of information retrieval, for comparison, where P@10 measures the correct
answers in top 10 cases. In our analysis, we measured P@5, P@10, P@15, and P@20; the results are
shown in Tables 6 and 7. Table 6 shows the comparison results among the ranking functions in the
case that type 1 is the true case and type 2 and 3 are false case whereas in Table 7, type 1 and 2 are the
true case and type 3 is the false case. The ranking of extracted 20 paths by P@K shows whether type
1 paths (or type 1 and 2) are ranked high, which implies the ranking algorithm properly predicts the
interesting, meaningful paths.
Genes 2019, 10, 159                                                                                       10 of 17

             Table 6. Comparison of precision at n (where type 1 = true and both type 2 and 3 = false).

                        Path Type      Co-Occurrence      COALS        UMLS       Proposed
                          P@5               0.20            0.40         0.00        0.60
                          P@10              0.30            0.50         0.20        0.60
                          P@15              0.40            0.40         0.33        0.40
                          P@20              0.30            0.30         0.30        0.30
                            COALS: please define; UMLS: Unified Medical Language System

             Table 7. Comparison of precision at n (where both type 1 and 2 = true and type 3 = false).

                       PATH TYPE       Co-Occurrence      COALS        UMLS       Proposed
                          P@5               0.80            0.60         0.40        0.80
                          P@10              0.70            0.90         0.50        0.80
                          P@15              0.73            0.67         0.67        0.73
                          P@20              0.65            0.65         0.65        0.65

     When type 1 is the only true case, the proposed method outperforms the other three methods by
18.75% to 128.92% for P@5-P@20. The second best performance was achieved by COALS, while UMLS
performed the worst. The proposed algorithm was particularly outstanding at top 5, and this result is
encouraging in cases where researchers want to investigate only a handful of the resulting hypotheses.
     When both type 1 and 2 are assumed to be true, the proposed method again outperforms the
other three methods, this time by 3.47% to 34.23%. The second best performance was achieved by the
co-occurrence method, while UMLS again did the worst.

3.3. Example of Literature Analysis
    One example for a type 1 path in which the entities are related and the verbs are correctly
connected is path 7. Path 7 shows the connection between the drug sorafenib and the side effect
dyspepsia. Figure 4 is conceptual model of path 7 that notes the evidence for each entity pair and verb.

3.3.1. Drug-Protein Connection: Sorafenib (Inhibit, Block) p38
     Sorafenib is a kinase inhibitor drug that is used to treat primary kidney cancer and advanced
primary liver cancer [32,33]. Uncontrolled growth in many cancers is due to a defect in the
Ras-Raf-MEK-ERK path, also known as the MAP/ERK path [34]. Sorafenib acts as an inhibitor
for several tyrosine protein kinases, such as VEGFR, PDGFR, and Raf family kinases, resulting in
the suppression of tumor growth [35]. Researchers have also shown that sorafenib can inhibit the
activation of the MAP kinase p38 by a marked decrease in p38 phosphorylation, without affecting total
protein levels [36,37]. These findings support the connection between sorafenib and p38, which are
linked by the verbs “inhibit” and “block.”

3.3.2. Protein-Protein Connection: p38 (Inhibit) g17
     P38 mitogen-activated protein kinases are one of the main subgroups of the MAP kinases that play
a vital role in signal transduction, cell differentiation, apoptosis, and senescence [38–41]. Gastrin-17
(G-17), also known as little gastrin 1, is a form of the protein hormone gastrin that is secreted by
the intestine [42]. Gastrin is produced in the G cells of the duodenum and in the pyloric antrum of
the stomach, and is released in response to certain stimuli such as hypercalcemia (elevated levels of
calcium in the blood) [43,44]. Gastrin stimulates hydrochloric acid/gastric acid secretion by inducing
histamine release from ECL cells, functioning as a central regulator for gastric acid secretion [45].
In our study, we combined gastrin-17 with gastrin, which also exists as gastrin-34 and gastrin-14, into a
single entity, as all three form of gastrin are produced in the G cells and functions similarly.
Genes 2018, 9, x FOR PEER REVIEW                                                                       11 of 17

Genes 2019, 10, 159                                                                                        11 of 17

                  Figure 4. Extracted drug-protein–side effect paths for sorafenib and dyspepsia.

3.3.3. Protein-Protein Connection: p38 (Inhibit) Gastrin
                      Figure 4. Extracted drug-protein–side effect paths for sorafenib and dyspepsia.
     We could not find studies directly connecting from p38 to gastrin, although there were studies
directly connecting from gastrin to p38. We searched for other factors that can connect from p38 to
Genes 2019, 10, 159                                                                                  12 of 17

gastrin and found NF-κB. NF-κB is a protein complex that regulates transcription, cytokine production,
and cell survival and is involved in multiple cellular responses [46,47]. Its transcriptional activation was
found to be regulated by the p38 MAP kinase activity [48–50], which upregulates NF-κB expression
through RelA phosphorylation during stretch-induced myogenesis [50]. NF-κB activity was also
induced in C2C12 cells by the activation of p38 [51]. IL1B-activated NF-κB downregulated gastrin,
and this downregulation occurred both in the presence and absence of IL1B [52,53]. The ectopic
expression of the p65 subunit of NF-κB in AGS cells resulted in about nine fold reduction in gastrin
levels, suggesting that gastrin is negatively regulated by NF-κB [53]. These findings suggest a
connection between p38 and gastrin by way of NF-κB, in which activation of p38 MAP kinase
upregulates NF-κB, which represses the transcription of gastrin.
     Another factor that can connect p38 to gastrin is calcium. Osteoclasts are a type of bone cell
that breaks down bone tissue, a process critical for bone maintenance, repair, and remodeling [54].
Previous research has found that p38 MAP kinase signaling plays a crucial role in PTHrP-induced
osteoclastic bone resorption, in which osteoclasts break down bone and result in the transfer of calcium
from bone fluid to the blood [55,56]. FR167653, an inhibitor of p38 MAP kinase, was found to inhibit
PTHrP-induced osteoclastogenesis in vitro and PTHrP-induced bone resorption in vivo [56]. Studies
also show that bone resorption induced by IL-1 and TNF is mediated by p38 MAP kinase and that
p38 activity enhances osteoclast maturation and bone resorption in myeloma [57,58]. These findings
suggest that p38 MAP kinase activity plays a crucial role in osteoclast maturation and bone resorption,
and thereby can regulate calcium levels in the blood. As mentioned, gastrin is released in response
to hypercalcemia (an elevated calcium level in the blood), suggesting that p38 can regulate gastrin
through calcium. However, more study is needed to understand exactly how p38 regulates calcium.
     Through NF-κB, we can support the connection of p38 to gastrin by the verb inhibit; p38 MAP
kinase up regulates NF-κB resulting in the inhibition of gastrin. Through calcium, we can support
the connection between p38 and gastrin, but cannot support the verb inhibit. However, through
our literature research, we have found a high correlation between p38 and gastrin through calcium,
suggesting that calcium is likely to be another mediator connecting p38 and gastrin.

3.3.4. Protein-Side Effect Connection: Gastrin (Associate) Dyspepsia
     Dyspepsia, also known as indigestion, is a condition in which digestion is impaired. Dyspepsia
is highly related to gastrin, as gastrin is a key regulator for the secretion of gastric acid, a digestive
fluid formed in the stomach [45]. Dyspepsia can be caused by gastroesophageal reflux disease (GERD),
a condition in which stomach acid comes up from the stomach into the esophagus and causes mucosal
damage. These findings support the connection between gastrin and dyspepsia by the verb associate.
     In short, we suggest dyspepsia as a side effect for sorafenib: sorafenib inhibits p38, thereby
inducinggastrin, which results in dyspepsia. Through our method, we suggest a mechanism for how
sorafenib can cause dyspepsia. For other analyses of type 2 and 3 paths, refer to the Supplementary
Data (Figures S1 and S2, Table S1 and Data S1).

3.4. Direct Link between Drug and Side Effect
     The list of the direct links between drug and side effect was extracted from the collected data,
and we provided it along with total frequency of drugs and side effects co-occurred in the same
abstract in Table 8.
     In Table 8, each drug and side effect is accompanied with Concept Unique Identifier (CUI)
provided in the UMLS. We extracted total 827 unique links provided in Appendix A. The most
frequently co-occurred pairs are Anti-inflammatories and Cachexia and Antioxidants and Cachexia
whose the total number of co-occurrences is 17,653.
Genes 2019, 10, 159                                                                                      13 of 17

                            Table 8. The top 20 direct links between drug and side effect.

                           Drug|CUI                          Side Effect|CUI                 Frequency
              ANTI-INFLAMMATORIES|C0003209                CACHEXIA|C0006625                   17,653
                 ANTIOXIDANTS|C0003402                    CACHEXIA|C0006625                   17,653
                       NSAIDS|C0003211                        LBP|C0024031                     7600
                         PCT|C0032452                         LBP|C0024031                     7296
                       SR59230A|C0386264                  CACHEXIA|C0006625                    3336
                         PD|C0030230                       FATIGUE|C0015672                    3234
                        IRON|C0302583                      FATIGUE|C0015672                    3234
                      IBUPROFEN|C0020740                  CACHEXIA|C0006625                    3197
                         TG|C0039902                      CACHEXIA|C0006625                    2919
                        DIET|C0012155                     CACHEXIA|C0006625                    2780
                      ANTIGEN|C0003320                    CACHEXIA|C0006625                    2502
                      GEFITINIB|C1122962                   FATIGUE|C0015672                    2464
                         TCC|C0077072                      FATIGUE|C0015672                    2310
                         SER|C0036720                     CACHEXIA|C0006625                    2224
                      ERLOTINIB|C1135135                   FATIGUE|C0015672                    2156
                         DTC|C0012194                      FATIGUE|C0015672                    2002
                       CNI-1493|C0384938                   FATIGUE|C0015672                    1848
                      PROTONS|C0033727                     FATIGUE|C0015672                    1848
                      HYDROGEN|C0020275                    FATIGUE|C0015672                    1848

4. Discussion

Advantage and Significance
     Although the side effects of drugs are reported in clinical trials, there are few studies that attempt
to explain why. However, there are many studies on the target paths of cancer drugs. Therefore in our
research we used text mining to identify the pathways between drugs and side effects. For example,
dyspepsia is a known side effect for Sorafenib. However, the entities connecting Sorafenib and
dyspepsia were not well known. Through our research, we suggest p38 and gastrin as entities that
can connect Sorafenib to its known side effect, dyspepsia. In situations where biological lab research
has its limitations, such methods can provide a path between drug and side effect with little time
and resources.
     In addition, we can suggest a meaningful hypothesis for further research into a drug and its side
effects by using our suggested ranking system. Although there could be many paths by which the side
effect may occur, our system provides an opportunity to preferentially overview plausible paths.
     Lastly, we think that our methodology could provide a more effective prescription of drug.
Currently when prescribing drugs, side effects are considered, and the prescription of drugs with
harsh side effect is often avoided. By extracting the path between drug and side effect with verbs,
our system can help doctors interpret the underlying path. For example, as shown above, p38 and
gastrin are related entities that connect Sorafenib to dyspepsia. Through our research, we suggest
that Sorafenib may cause dyspepsia by inhibiting p38, thereby inducing gastrin, which may result in
dyspepsia. With this information, doctors can assess a patient’s risk of suffering a given side effect,
and therefore, more effectively prescribe the drug in question.
Genes 2019, 10, 159                                                                                            14 of 17

5. Conclusions
      We used text mining to identify possible paths between drugs and side effects by analyzing
2,379,349 research paper abstracts. By extracting the relation between entities from the abstracts’ free
text, we suggest detailed mechanisms connecting drugs and side effects.
      To make the results suggested by the proposed approach more reliable, we need to apply methods
that can increase the accuracy of suggested specific processes such as NER, RE, and path analysis,
including a new path ranking algorithm, and also a develop a process for analyzing error cases.
Moreover, a method that can consider conditions and situations that affects the relation will be needed
in our future work. In addition, since we used a limited number of oncogenes as search queries,
it would be desirable to include more comprehensive oncogenes to search queries.
      Our proposed framework is not limited to drug–side effect relations, however, and could be
applied in various circumstances. For this, it requires first defining a detailed conceptual model;
for example, what entity types could be contained in the path. Then, a detailed process such as NER,
RE, and path analysis are executed to extract potential entities. Through these pipelines, our suggested
method may give meaningful results for biomedical researchers in various situations such as
protein-protein interaction (PPI). For PPI, we are able to integrate widely used PPI databases such as
Mammalian Protein-Protein Interaction Database (http://mips.helmholtz-muenchen.de/proj/ppi/)
and BioGRID (http://thebiogrid.org/) for PPI extraction.

Supplementary Materials: The following are available online at http://www.mdpi.com/2073-4425/10/2/159/s1,
Figure S1: Path1 (Type 2) and Path 2 (Type 3), Figure S2: Path 15 (Type 2), Table S1: Ranking function result,
File S1: Further literature analysis.
Author Contributions: Conceptualization, M.S. and S.H.B.; Methodology, M.S., G.E.H. and J.-H.L.; Software,
G.E.H.; Validation, S.H.B., G.E.H. and J.-H.L.; Formal Analysis, S.H.B.; Investigation, M.S. and S.H.B.; Resources,
G.E.H.; Data Curation, S.H.B. and G.E.H.; Writing—Original Draft Preparation, M.S., S.H.B. and G.E.H.;
Writing—Review & Editing, M.S. and J.-H.L.; Supervision, M.S.; Project Administration, M.S.; Funding
Acquisition, J.-H.L.
Acknowledgments: This research was supported by Next-Generation Information Computing Development
Program through the National Research Foundation of Korea (NRF), funded by the Ministry of Science,
ICT (No.NRF-2017M3C4A7065887).
Conflicts of Interest: The authors declare no conflict of interest.

References
1.    Wood, A.J.; Evans, W.E.; McLeod, H.L. Pharmacogenomics—Drug disposition, drug targets, and side effects.
      N. Engl. J. Med. 2003, 348, 538–549.
2.    Kuhn, M.; Campillos, M.; Letunic, I.; Jensen, L.J.; Bork, P. A side effect resource to capture phenotypic effects
      of drugs. Mol. Syst. Biol. 2010, 6, 343. [CrossRef] [PubMed]
3.    Gurulingappa, H.; Fluck, J.; Hofmann-Apitius, M.; Toldo, L. Identification of adverse drug event
      assertive sentences in medical case reports. In Proceedings of the First International Workshop on
      Knowledge Discovery and Health Care Management (KD-HCM), European Conference on Machine
      Learning and Principles and Practice of Knowledge Discovery in Databases (ECML PKDD), Athens, Greece,
      5–9 September 2011; pp. 16–27.
4.    Atias, N.; Sharan, R. An algorithmic framework for predicting side effects of drugs. J. Comput. Biol. 2011, 18,
      207–218. [CrossRef] [PubMed]
5.    Huang, L.-C.; Wu, X.; Chen, J.Y. Predicting adverse side effects of drugs. BMC Genom. 2011, 12, 1. [CrossRef]
      [PubMed]
6.    Chen, B.; Ding, Y.; Wild, D.J. Assessing drug target association using semantic linked data. PLoS Comput. Biol.
      2012, 8, e1002574. [CrossRef] [PubMed]
7.    Lounkine, E.; Keiser, M.J.; Whitebread, S.; Mikhailov, D.; Hamon, J.; Jenkins, J.L.; Lavan, P.; Weber, E.;
      Doak, A.K.; Côté, S. Large-scale prediction and testing of drug activity on side-effect targets. Nature 2012,
      486, 361–367. [CrossRef] [PubMed]
Genes 2019, 10, 159                                                                                           15 of 17

8.    Rebholz-Schuhmann, D.; Oellrich, A.; Hoehndorf, R. Text-mining solutions for biomedical research: Enabling
      integrative biology. Nat. Rev. Genet. 2012, 13, 829–839. [CrossRef] [PubMed]
9.    Sohn, S.; Kocher, J.P.; Chute, C.G.; Savova, G.K. Drug side effect extraction from clinical narratives of
      psychiatry and psychology patients. J. Am. Med. Inform. Assoc. 2011, 18 (Suppl. 1), i144–i149. [CrossRef]
      [PubMed]
10.   Thompson, P.; McNaught, J.; Montemagni, S.; Calzolari, N.; Del Gratta, R.; Lee, V.; Marchi, S.; Monachini, M.;
      Pezik, P.; Quochi, V.; et al. The BioLexicon: A large-scale terminological resource for biomedical text mining.
      BMC Bioinform. 2011, 12, 397. [CrossRef] [PubMed]
11.   Pauwels, E.; Stoven, V.; Yamanishi, Y. Predicting drug side-effect profiles: A chemical fragment-based
      approach. BMC Bioinform. 2011, 12, 169. [CrossRef] [PubMed]
12.   Hopkins, A.L. Network pharmacology: The next paradigm in drug discovery. Nat. Chem. Biol. 2008, 4,
      682–690. [CrossRef] [PubMed]
13.   Berger, S.I.; Iyengar, R. Network analyses in systems pharmacology. Bioinformatics 2009, 25, 2466–2472.
      [CrossRef] [PubMed]
14.   Li, J.; Zhu, X.; Chen, J.Y. Building disease-specific drug-protein connectivity maps from molecular interaction
      networks and PubMed abstracts. PLoS Comput. Biol. 2009, 5, e1000450. [CrossRef] [PubMed]
15.   Barabási, A.-L.; Gulbahce, N.; Loscalzo, J. Network medicine: A network-based approach to human disease.
      Nat. Rev. Genet. 2011, 12, 56–68. [CrossRef] [PubMed]
16.   Cami, A.; Arnold, A.; Manzi, S.; Reis, B. Predicting adverse drug events using pharmacological network
      models. Sci. Transl. Med. 2011, 3, 114ra127. [CrossRef] [PubMed]
17.   Lee, S.; Lee, K.H.; Song, M.; Lee, D. Building the process-drug–side effect network to discover the relationship
      between biological Processes and side effects. BMC Bioinform. 2011, 12, 1. [CrossRef] [PubMed]
18.   Xu, R.; Wang, Q. Automatic construction of a large-scale and accurate drug-side-effect association knowledge
      base from biomedical literature. J. Biomed. Inform. 2014, 51, 191–199. [CrossRef] [PubMed]
19.   Sharma, P.; Senthilkumar, R.; Brahmachari, V.; Sundaramoorthy, E.; Mahajan, A.; Sharma, A.; Sengupta, S.
      Mining literature for a comprehensive pathway analysis: A case study for retrieval of homocysteine related
      genes for genetic and epigenetic studies. Lipids Health Dis. 2006, 5, 1. [CrossRef] [PubMed]
20.   Campillos, M.; Kuhn, M.; Gavin, A.-C.; Jensen, L.J.; Bork, P. Drug target identification using side-effect
      similarity. Science 2008, 321, 263–266. [CrossRef] [PubMed]
21.   Eleftherohorinou, H.; Wright, V.; Hoggart, C.; Hartikainen, A.-L.; Jarvelin, M.-R.; Balding, D.; Coin, L.;
      Levin, M. Pathway analysis of GWAS provides new insights into genetic susceptibility to 3 inflammatory
      diseases. PLoS ONE 2009, 4, e8068. [CrossRef] [PubMed]
22.   He, B.; Tang, J.; Ding, Y.; Wang, H.; Sun, Y.; Shin, J.H.; Chen, B.; Moorthy, G.; Qiu, J.; Desai, P. Mining
      relational paths in integrated biomedical data. PLoS ONE 2011, 6, e27506. [CrossRef] [PubMed]
23.   Li, B.-Q.; Huang, T.; Liu, L.; Cai, Y.-D.; Chou, K.-C. Identification of colorectal cancer related genes with
      mRMR and shortest path in protein-protein interaction network. PLoS ONE 2012, 7, e33393. [CrossRef]
      [PubMed]
24.   Ramanan, V.K.; Shen, L.; Moore, J.H.; Saykin, A.J. Pathway analysis of genomic data: Concepts, methods,
      and prospects for future development. Trends Genet. 2012, 28, 323–332. [CrossRef] [PubMed]
25.   Rohde, D.L.; Gonnerman, L.M.; Plaut, D.C. An improved model of semantic similarity based on lexical
      co-occurrence. Commun. ACM 2006, 8, 627–633.
26.   Jurgens, D.; Stevens, K. The S-Space package: An open source package for word space models. In Proceedings
      of the ACL 2010 System Demonstrations, Uppsala, Sweden, 13 July 2013; pp. 30–35.
27.   McInnes, B.T.; Pedersen, T.; Pakhomov, S.V. UMLS-Interface and UMLS-Similarity: Open source software for
      measuring paths and semantic similarity. In Proceedings of the AMIA Annual Symposium Proceedings,
      San Francisco, CA, USA, 14–18 November 2009; p. 431.
28.   Garla, V.N.; Brandt, C. Semantic similarity in the biomedical domain: An evaluation across knowledge
      sources. BMC Bioinform. 2012, 13, 261. [CrossRef] [PubMed]
29.   Manning, C.D.; Surdeanu, M.; Bauer, J.; Finkel, J.R.; Bethard, S.; McClosky, D. The Stanford CoreNLP Natural
      Language Processing Toolkit. In Proceedings of the ACL (System Demonstrations), Baltimore, MD, USA,
      22–27 June 2014; pp. 55–60.
30.   Song, M.; Kim, W.C.; Lee, D.; Heo, G.E.; Kang, K.Y. PKDE4J: Entity and relation extraction for public
      knowledge discovery. J. Biomed. Inform. 2015, 57, 320–332. [CrossRef] [PubMed]
You can also read